We\u2019re pleased to share our recent work published in\u00a0Expert Systems with Applications (Elsevier),\u00a0on a\u00a0Decision Support System (DSS)\u00a0that models industrial water networks as process networks consisting of:
\u00a0\u2022\u00a0\u00a0Freshwater and wastewater sources
\u00a0\u2022\u00a0\u00a0Treatment processes
\u00a0\u2022\u00a0\u00a0Connection links with capacity, quality, and reduction-rate attributes
The DSS is powered by a compact mathematical model automatically generated from operational data through our\u00a0optimisation engine -optEngine-\u00a0optimising cost, energy, freshwater intake, and wastewater treatment.
Crucially, the system supports both design-time and operational decision-making, enabling:
\u00a0\u2022\u00a0\u00a0Evaluation of alternative network design investments
\u00a0\u2022\u00a0\u00a0Day-to-day optimisation of water flows
\u00a0\u2022\u00a0\u00a0Rapid response to unexpected operational events
Validated on large-scale industrial case studies, the framework demonstrates how advanced optimisation can be translated into practical, real-time decision support for complex water networks.
<\/p>\n\n\n\n
Optiscale: a decision support system for optimised industrial water management<\/strong><\/p>\n\n\n\n Industrial water networks are highly complex. They involve intricate process topologies, multiple water quality constraints, and frequent disruptions such as valve malfunctions, pipeline blockages, and demand fluctuations. Managing freshwater intake and reuse efficiently under these conditions remains a major industrial challenge.<\/p>\n\n\n\n \u00a0OptEngine in action<\/strong><\/p>\n\n\n\n Our optimisation engine -OptEngine- operates daily or on-demand, translating network data into mathematical programming models and proposing optimised (waste)water flow configurations across the network.<\/p>\n\n\n\n \ufe0f\u00a0Performance highlights<\/strong><\/p>\n\n\n\n \u2022 Near-optimal solutions delivered within seconds<\/p>\n\n\n\n \u2022 Tested on networks with 5\u201315 components and 5\u201310 quality parameters<\/p>\n\n\n\n \u2022 Suitable for real-time operational support<\/p>\n\n\n\n \u00a0Validated on real industrial case studies<\/p>\n\n\n\n The framework was validated using data inspired by three large-scale industrial applications, including:<\/p>\n\n\n\n \u00b7 An oil refinery (T\u00fcpra\u015f, Izmit, Turkey)<\/p>\n\n\n\n \u00b7 Two chemical industries (DOW, Bohlen, Germany & Solvay, Livorno, Italy)<\/p>\n\n\n\n \u00a0Measured impact<\/p>\n\n\n\n \u00b7 Water reuse up to 90% (~2.3 mcm\/year)<\/p>\n\n\n\n \u00b7 Freshwater intake reduction up to 18% (~1.2 mcm\/year)<\/p>\n\n\n\n \u00b7 Support for selecting the most cost-effective network design investments<\/p>\n\n\n\n Why this matters<\/strong><\/p>\n\n\n\n Unlike traditional, case-specific optimisation solutions, this work delivers a generalised, reusable decision-support framework\u2014bridging advanced optimisation theory with practical industrial water management. From design to operations, this is how decision support systems enable resilient, efficient, and sustainable industrial water networks\u2014without compromising production.<\/p>\n\n\n\n